The Effects of Monopoly in Information Technology on...

The Effects of Monopoly in Information Technology on Democracy: A New Perspective on Microsoft's Monopoly

Departmental Honors Thesis

Department of Government and Politics Department of Computer Science

University of Maryland, College Park

David Thaw

Advisors: Dr. Wayne McIntosh and Dr. Ashok Agrawala

May 12, 2003

David Thaw University of Maryland, College Park Departmental Honors Thesis of 57 Section I: Introduction

The concept of free exchange of information and ideas is a central tenet of

democratic theory. This exchange is crucial both to the deliberative process necessary for

democracy and to the elements of personal liberty and self-actualization that are at the

heart of democracy. For individuals to have the ability to make choices freely, they must

have access to information relevant to their choices. Raw information, however, is

generally too complex and/or diffuse to be processed in the time frame that individuals

need for decision-making purposes. To be useful, information must be aggregated,

sorted, and analyzed. In short, it must be processed according to individual needs and

desires. Ideally, economic theory would suggest that competitive markets for this type of

“pre-processed” information would result, and market economic principles would ensure

that this “product” would be produced according to the demands of its consumers.

Information, particularly in modern times, requires a complex delivery system.

As individuals demand increasingly complex and comprehensive information for their

decision-making processes, information providers seek to develop better ways of

organizing and delivering this information to their consumers. Indeed, recent years have

shown a staggering growth in both the complexity and usage of modern information

technologies including Personal Computers (PCs), mobile computing devices, wireless

communications, and the Internet. Advances in computing technology, thus, have

seemingly enabled information providers to better meet their consumers’ demands.

While several other technological alternatives exist in widespread use, in recent

years computer-based technologies (broadly understood) have emerged as a leading

facility of information flow and communication. In particular, PCs and their myriad

David Thaw University of Maryland, College Park Departmental Honors Thesis of 57 communication software, PDAs, and cellular phones have dominated new equipment

purchases for information management and exchange. At the heart of these systems is a

technology known as the operating system, that software which controls the operation of

the mechanical and electronic components of computing devices and provides a common

platform on which task-oriented software applications reside. Were a private entity to

become the primary (monopolistic) provider of operating systems, such an entity could

obtain the ability to significantly influence the flow of information. Over the course of

the past several years, it has become clear that such an entity has emerged – Microsoft

Corporation.

Section II: Principles of Democratic Society

Recent developments in information technology and the importance of the free

exchange of information to democratic society suggest the importance of examining the

implications of the integration of these new technologies into society. Those aspects of a

democratic society most affected by the flow of information provide the starting point for

this investigation. This paper focuses primarily on two such aspects: 1) individuality and

the process of self-actualization, and 2) the development and maintenance of effective

democratic participation. It is through these two metrics that I will evaluate the potential

impact of modern information technologies and attempt to prove that the potential exists

for these principles to be aversely affected if such technologies continue to develop

unchecked.

Information, broadly understood, is perhaps the most crucial element of individual

decision-making processes. When individuals are faced with a decision, whether it be

David Thaw University of Maryland, College Park Departmental Honors Thesis of 57 choosing a movie to view or a candidate to vote for in an election, they collect

information about the choices they are faced with and integrate that information into their

decision-making process. Similarly, when forming opinions, individuals collect

information which they then process and integrate into the formation of those opinions.

Individuality, Liberty, Freedom of Pursuit and Self-Actualization

J.S. Mill, in his essays On Liberty, argues that the concept of individuality, that

right of each person to freely interpret experience and information, to form opinions, and

– within appropriate forums – to express those opinions, is one of the essential elements

of human “well-being.” “[I]t is the privilege and proper condition of a human being,

arrived at the maturity of his faculties, to use and interpret experience in his own way”

(Mill 56). Mill’s argument is particularly important when taken in the context of modern

information delivery. The acquisition and processing of information is a critical element

of an individual’s experience. If, therefore, the whole of information available were to be

pre-processed by some other entity, that pre-processing might endanger the individual’s

right to “interpret [that information and] experience in his own way.” While “nobody

denies that people should be so taught and trained in youth as to know and benefit by the

ascertained results of human experience” (Mill 56), “[i]t is for [each individual] to find

out what part of recorded experience is properly applicable to [his or her] own

circumstances and character” (56). Furthermore, it is the right of each individual to

develop such circumstances and character, subject to reasonable guidelines necessary for

the preservation of individual rights, free of influence and coercion.

David Thaw University of Maryland, College Park Departmental Honors Thesis of 57 Effective Democratic Participation

The free exchange of information and ideas is also a critical element of the

democratic process and the culture of a democratic society. “Rights are among the

essential building blocks of a democratic process of government” (Dahl 48). Effective

participation, for example, requires that citizens “[p]ossess a right to participate, a right to

express their views on political matters, to hear what other citizens have to say, [and] a

right to discuss political matters with other citizens… citizens must have a right to

investigate alternatives [and] a right to participate in deciding how and what should go on

the agenda” (Dahl 48-49). The flow of information is essential to the realization of each

of these rights, particularly in large modern democracies. Increasing population size,

interdependence of geographically separate regions, and the process of globalization

requires that an individual have access to information about a much broader range of

issues to effectively participate than in earlier times in democratic history.

Section III: The Importance of Information Exchange to Democratic Society

The distribution of information requires some form of mechanism to support it.

Assuming that there exist common forms of expression for information (i.e., languages,

visual representations, etc.) there must then exist mechanisms that can be used to

disseminate these representations of information to consumers who desire it. Historical

analysis of these mechanisms reveals that these mechanisms have continued to develop

over time, and significant advances in these mechanisms often coincided with significant

scientific, industrial, economic, and/or cultural advances (e.g., the development of the

printing press and the spread of religion and culture throughout Europe, the development

David Thaw University of Maryland, College Park Departmental Honors Thesis of 57 of the US postal system and the early economic development of the United States). In

each of these cases, the development and integration of new communications and

information technologies had significant developmental consequences and influence over

society and the economy. Likewise, with the introduction of each new information

technology, the variety of information sought by consumers and the number of

information producers increased. As each new technology provided for a more complex

delivery system, science, culture, and the economy advanced adopting these systems and

developmentally benefiting from them.

An Example – The Printing Press and Postal Systems

The printing press and the postal system are excellent examples of the influence

that information distribution mechanisms can have on society and democratic processes.

As Francis Grund stated, “the postal system had, in conjunction with the periodical press,

done more to change ‘the face of the world’ than ‘half a million of philosophers’ or ‘the

bayonets of all the nations of Europe’” (John 12). Unlike person-to-person contact, such

as debate and public speaking, these systems’ influence was inherent in the mechanisms

themselves rather than the information they conveyed. As John describes, “[a] given

newspaper article, Grund conceded, might not exert as great an influence as a particularly

moving speech. But if the message were repeated again and again, its effect was sure to

be far more profound” (John 13). These mechanisms enabled messages to be “repeated

again and again” throughout the United States, reaching more people across a broader

geographical range than any public speaker could.

David Thaw University of Maryland, College Park Departmental Honors Thesis of 57

The influence of information distribution mechanisms was reflected not only in

their ability to influence individuals directly, but also in their ability to facilitate

democracy. In John’s analysis of the U.S. Postal System as an agent of change, he

examines the work of Eli Bowen, a nineteenth-century journalist and postal clerk.

Bowen compared public life in the United States to that of the Greek and Roman empires.

In particular, he focused on the failure of these empires to develop a form of democracy

that included those individuals who did not reside in the foci of government found in the

cities. “’We hear nothing,’ Bowen observed, of… that class of landowning citizen-

farmers who could participate in public life without ever leaving the farm” (John 14).

According to John, Bowen argued that “[s]hould a Greek or Roman citizen-farmer wish

to take part in public affairs, he had no choice but to leave his farm and move to the city,

since he needed a good deal of information to participate effectively in public affairs,

while this information was necessarily restricted to those individuals who lived in close

physical proximity to the seat of power” (John 14). In contrast, for the United States,

“[t]hanks to the postal system, the citizen-farmer had no trouble securing access to a

steady flow of information on public affairs, making it possible for him to participate in

national politics without leaving the farm” (John 15). These observations were shared by

Alexis de Tocqueville, who calculated that in the mid-eighteenth century, based on postal

revenue, “the average inhabitant of Michigan territory received a greater volume of

nonlocal information than the average inhabitant of the Department du Nord,” a bustling

commercial center at the heart of France (John 1).

These observations demonstrate the importance of information and its distribution

to a democratic society. The success of the United States as the first geographically

David Thaw University of Maryland, College Park Departmental Honors Thesis of 57 wide-scale democracy of modern times was significantly incumbent upon the use of these

and future “information technologies.” The postal system and the printing press were

crucial in “transform[ing the United States] from a confederation of separate states into

‘one great neighborhood’” (John 13). Having established the importance of information

exchange and distribution to democratic society, I then move on to examine the

conditions for the production and distribution of information.

Section IV: Information Markets

Raw information is generally too complex and/or diffuse to be processed in the

time frame that individuals need for decision-making purposes. To be useful, therefore,

information must be aggregated, sorted, and analyzed. In short, it must be processed

according to individual needs and desires. Historically, information technologies have

adapted to meet this requirement by providing palettes of choices from which individuals

can choose. The printing press, for example, allowed the manufacture of periodicals that

were tailored to various needs and desires. Similarly, broadcast radio has a variety of

stations and broadcast television has a variety of channels. Thus the mechanisms for

processing and distributing information – “information technologies” – become a critical

link in the system of information exchange and as such have a meaningful influence on

the content of that exchange.

Individuals comprise the primary consumers of information. Individuals use

information for a variety of purposes, including decision-making, education, and

entertainment. As discussed earlier, Mill argues that the right of individuals to freely

interpret experience and information is an essential component of individual liberty.

David Thaw University of Maryland, College Park Departmental Honors Thesis of 57 Mill’s discussion leads into the concept of “self-actualization,” the idea that individuals

in a democratic society have a right to freely develop themselves (within necessary

guidelines for the maintenance of equal rights). This development includes an

individual’s right to educate and entertain him/herself as he/she sees fit, and the right to

freely make decisions based on information without hidden bias.

Information Biases – The Production Problem

The concept of “hidden bias” derives from the idea that the same information can

be analyzed, aggregated, interpreted, and packaged in more than one fashion and from

more than one standpoint. Two scientists, for example, may examine the same set of

laboratory observations yet come to completely different conclusions. Examples of this

phenomenon are extremely common in American society, and can be witnessed by

perusing almost the science section of any newspaper of record on almost any given day.

These biases then affect the information end-product by analyzing, interpreting, and

packaging the raw data that leads to information delivery to consumers in a manner

consistent with the particular bias.

It is likely impossible, under the current state of human affairs, to expect that any

or all sources and distribution systems of information will be unaffected by such biases. I

examine two potential ways in which these biases can manifest themselves: 1) the

information delivery systems, and 2) the producers of the end-point information delivered

by their respective distribution systems. The former refers to the case in which the

owners of the information delivery technology prefer a certain set of information sources

David Thaw University of Maryland, College Park Departmental Honors Thesis of 57 and/or opinions and provide producers favorable access to their delivery media. The

latter refers to the case, an example of which is given above with “the two scientists,” in

which one producer interprets data differently from another. It is important to note that

these two manifestations are not clearly distinct from one another and often operate

together. In the case of newspapers, for example, one paper with a particular set of

viewpoints may not only prefer certain sources/producers, but may also seek out such

individuals and entities and employ them to assemble the information package the paper

seeks to distribute. A classic example of this phenomenon is the New York City

Metropolitan area, where the New York Times (a comparatively liberal paper) competes

with the Wall Street Journal (a comparatively conservative paper).1

Rather, therefore, than examining these conditions separately, I will examine

them together with the focus on the results affecting how delivery system owners will and

should behave. This analysis begins by returning to individuals the primary consumers.

Taking a standard market-economic approach, and assuming that the consumer-

individuals in question are members of and wish to promote a democratic society,2 I

examine how the market will manifest and respond to individuals’ preferences in order to

1 It should be noted that this example is perhaps an unfair generalization of these papers, which are aimed not only at ideologically different audiences, but also at audiences with different substantive interests. The Times presents more of a generalist approach to news reporting, whereas the Journal focuses more on fiscal news. 2 This is perhaps an unfair assumption about the majority of people in modern (American) society. Social Psychologists debate on whether altruism actually exists (Myers 474-485), and some argue that it is really a manifestation of social exchange theory, or “the theory that human interactions are transactions that aim to maximize one’s rewards and minimize one’s costs” (Myers 474). Nevertheless, the argument can proceed based on the assumption that individuals will seek these democratic goals because in the end it will all be to their mutual benefit.

David Thaw University of Maryland, College Park Departmental Honors Thesis of 57 produce outcomes desirable to democracy under the conditions described in Section II.

(Baker 2002 63-95).

Individuals As Information Consumers

“The standard claim on behalf of the market is that it responds to people’s real

preferences; it provides audiences with what they want” (Baker 2002 80). Baker rightly

argues that this proposition fails in the case of identifying consumer media preferences,

for “[e]ven if the existing distribution of wealth were a fully appropriate criterion for

weighting people’s preferences… the market does not properly identify people’s

preferences” (Baker 2002 81). It can be reasonably dispensed with that the existing

distribution of wealth is not “a fully appropriate criterion.” According to Dahl, the

democratic process entails (amongst other things) ‘[a] weak principle of equality: The

good of each person is entitled to equal consideration” (Dahl 1985 57). Combine this

with the proposition that “ownership and control contribute to the creation of great

differences among citizens in wealth, income, status, skills, information, control over

information and propaganda… [and] after all due qualifications have been made,

differences like these help in turn to generate significant inequalities among citizens in

their capacities and opportunities for participating as political equals in governing the

state” (Dahl 1985 54). It is general knowledge that a small top percentile of people in the

United States controls the largest percentage of wealth (which includes corporate

ownership and control). From these statements then it follows, simply put, that all

citizens do not have equal “capacities and opportunities” to participate as political equals,

or, applied to Baker’s argument, to express their preferences. For just as Dahl’s political

David Thaw University of Maryland, College Park Departmental Honors Thesis of 57 inequalities present concern for effective citizen participation and consideration, the

economic inequalities present in the United States present obstacles to equal preference

identification in media markets predicated upon the idea that each citizen who wishes to

express a preference should be entitled to a (relatively) consistent recognition of that

preference (i.e., the concept of pluralism). In essence, therefore, those with greater

economic capital are far better able to express their preferences and receive the media and

information they desire.3

The Market As A Mechanism For Information Production and Distribution

Given the importance of information distribution mechanisms, media producers,

and the products they create to a democratic society, it is therefore important to examine

what means of organization are necessary (and possibly even what means are desirable)

to maintain the principles of such a society. The market being the most prevalent

organizational mechanism in the United States, we turn to it as the default case.4 As

examined above, individual citizens are the primary consumers in this case. Yet these

consumers share a great deal of inequality, presenting concern about which preferences

will be represented. For this mechanism to be effective, therefore, individuals must have

3 There is a potential corollary argument from this proposing that, in the case of media choice, the preferences of economically underprivileged individuals are even further suppressed as a function of advertising. In short, those with greater economic advantage better express their preferences, those preferences are then translated into information products, which are then supported (“paid-for”) by advertising aimed at the broader (economically disadvantaged) audience. Assuming the nature of advertising is such that it is aimed at convincing individuals of their “need” to purchase a product/service, it follows that these individuals utility to express their preferences would be further restricted both by the preferences forced upon them via the advertising and (possibly) by the reduced utility (economic assets) as a result of their purchasing the products/services they might ordinarily not have. 4 An examination of comparative methods of information technology and media production organizations across different (democratic) nations would likely provide further useful insight into the issues discussed in this paper.

David Thaw University of Maryland, College Park Departmental Honors Thesis of 57 the ability to collectively express preferences in a manner that will both overcome

potential inequalities and that will affect meaningful change among information

technology and media content producers.

Under standard principles, “[m]arkets measure preferences only when a person

makes purchases” (Baker 2002 81). This presents a particular problem in the case of

media products, as Baker effectively identifies (2002 7-19). “Under certain purportedly

normal circumstances, the market provides firms with an incentive to produce and sell the

product as long as the product’s cost is less than the marginal price. The market thereby

leads to a preference-maximizing production and distribution” (2002 7). As Baker

correctly identifies, however, this is not the case with markets for media products, and, I

will argue, their associated information technologies.5 Media and communications

products are characterized by four aspects that differentiate them from other products

(e.g., toasters) to which this model effectively applies: 1) they have significant “public

good”6 aspects; 2) they produce significant externalities; 3) two very different purchasers

are involved in the transaction; and 4) there are a wide variety of different ways in which

individuals “value” these products (2002 8-14).

Baker’s analysis effectively shows how media and communications technologies

exhibit the “nonrivalrous use” quality of public goods (2002 8-10). I take this argument a

5 Baker’s economic analysis of media products (2002 9, 20-22) can be applied to the production of modern information technologies, particularly computer software (especially operating systems), as well. These products share the same problem that media products do of incurring significant “up-front” costs of production but being relatively easy to produce after the initial “copy” is created. 6 “[A]n item for which one person’s use of or benefit from the product does not affect its use by or benefit to another person” (Baker 2002 8). Other definitions of public goods require the good to be both non-rival and non-excludable; in this case I adopt Baker’s definition since certain information, communications, and media products are excludable as a function of the nature of their technology. Furthermore, recent efforts by American businesses (e.g., Microsoft’s advanced anti-piracy technology in Windows XP and Office XP) indicate an increasing trend towards excludability in potential public goods.

David Thaw University of Maryland, College Park Departmental Honors Thesis of 57 step further, and assert that it can be similarly applied to modern information

technologies, specifically, computer software products. Like Baker’s “media products,”

computer software is a good for which “one person’s use of or benefit from the product

does not affect its use by or benefit to another person” (Baker 2002 8).7 Person A using

Microsoft Windows XP Professional as the operating system for his/her computer does

not prevent, limit, or restrict Person B from doing so. Rather, it may even make it easier

for person B to do so since they will likely be better able to assist one another and

collaborate since they will have a common platform on which to work. Also, much like

telecommunications systems, the production of computer software incurs a significant

“up-front” cost. This cost encompasses the design and development of the software.

Once completed, however, production of additional copies of this software incurs

comparatively trivial costs. With telecommunications infrastructure, “[m]ultiple

consumers can use this infrastructure with no or very modest extra expense. To the

extent that adding an additional customer does not increase the cost of this infrastructure,

which is usually true until crowding requires larger lines or mains, this infrastructure

exists as a public good” (Baker 2000 8). Likewise, with computer software, multiple

users can acquire and use additional copies of it with little marginal cost. Furthermore,

the concern described in Footnote 7 is addressed by Baker’s attention to the case of

marginal cost being negligible until crowding requires additional infrastructure changes.

With computer software, the marginal cost of technical support and other post-

7 This is not strictly true as there are certain marginal costs associated with the next user’s use (e.g., technical support, update service request drain on servers, etc.). I would argue that these costs, however, are limited. It may, however, be a worthwhile exercise to attempt to characterize these costs to determine if they do affect the non-rival assertion. As of the time of this writing, however, I was unable to find sufficient data to support such a claim.

David Thaw University of Maryland, College Park Departmental Honors Thesis of 57 distribution services does not significantly increase the production cost until additional

“infrastructure” is required as a result of crowding.

As is the case with Baker’s communications and media products, “[t]his creates a

problem” (2002 8). In short, the average cost of each software copy – while decreasing

as a function of sales – is significantly higher than the marginal cost of each software

copy. The problem results from this disparity. “If the product is priced at its average

cost, some consumers will be unwilling (or unable) to pay that price, even though they

want [or need] the products and would be willing to pay the added cost created by their

usage. Charging the average cost [therefore] results in underproduction” (2002 9). To

correct this problem, the seller could charge the marginal cost, “as efficiency

considerations normally recommend” (2002 9), however this would produce insufficient

revenue as the bulk of the costs are incurred pre-production, not at marginal production.

As a result, insufficient incentives to produce would arise and innovation and

development would suffer. (Baker 2002 8-10)

Baker’s second criterion examines externalities created by media and

communications goods production. He essentially discusses the potential benefits to and

concerns for “a well-functioning democracy” (2002 10) created by the production and

consumption of these goods. As discussed above, the flow and exchange of information

is essential to such a society. Indeed, with modern information technologies such

externalities continue to exist. In the case of computer software, however, additional

externalities arise which Baker’s analysis does not address. In the case of computer

operating systems and office software, for example, the facilitation of collaboration and

interoperability (or lack thereof) is a key externality. If many computing systems are

David Thaw University of Maryland, College Park Departmental Honors Thesis of 57 running the same software, and as a result, are fully interoperable, then the users of those

systems will better be able to collaborate on projects and exchange information and ideas.

The cost of doing so will be reduced as compared to groups of users running different

software, where collaboration and information exchange will incur the added cost of

“porting” that information across platforms and/or software packages.8 The resultant

externalities come from the increased or decreased ability for groups to collaborate and

innovate. Take, for example, two researchers at different universities working on a

technical development project. If these two scientists are using the same technical

development software running on the same operating system platform, they will more

easily be able to collaborate and share their work and will be able to more quickly

complete their research. Assuming that the research is meaningful and important to

society, its more rapid completion results in a positive externality to those individuals that

will benefit from the research. Conversely, if they are not running the same software

and/or systems, and their research is accordingly slowed, a negative externality will result

for those that would have benefited from the research were it completed earlier.

Baker’s third criterion introduces the idea that individual citizens, while the

primary consumers in the transaction of information exchange, are not the only

purchasers involved in these transactions. Baker argues that advertisers comprise this

second category of purchasers – even though they do not necessarily “consume” the

products in the traditional sense – and that “[s]elling to both audiences and advertisers

8 It is important to note that these costs have likely reduced over time, as software has been developed for multiple operating systems platforms and communications infrastructure (specifically, the internet) has facilitated transfer across system platforms. These costs, however, have not been eliminated and I would suspect are likely still significant. This is another question for which I do not believe sufficient data exists and would be worthwhile to examine.

David Thaw University of Maryland, College Park Departmental Honors Thesis of 57 has especially significant consequences and adds special complexities” (2002 11). In the

case of media products, as Baker identifies, this is certainly the case (2002 11-12). In the

case of computer software, however, this trinary sales relationship takes on a different

form. While some computer software does incorporate advertising,9 for the most part

commercial software does not depend on advertising revenues. One potential exception

to this rule, however, is Microsoft’s operating system software. In short, the idea behind

this is that the operating system is the first thing a user sees when he/she buys a new

computer and the first thing he/she sees when turning on the computer each day. It

therefore affords the OS manufacturer the opportunity to “steer” users in the direction of

certain producers for things like Internet access, specific classes of software (e.g., by

providing a trial version of an office software package and then offering easy instructions

for how to order the full version), and other communications services such as email or

instant messaging. A further examination of this will be provided in Section VII.

Baker’s fourth criterion examines the manner in which consumers “value” media

and communications products. Individuals lack, he argues, sufficient means by which to

evaluate if they have really received what they “wanted” when selecting such products.

As Baker states, “[a consumer’s] own preferences do not her give a complete standard by

which to measure whether her purchases provide the right thing” (2002 12). Hence they

must fall back to more general “rules of thumb for guessing whether they are likely to get

[what they want]” (2002 13). Baker identifies two such “rules of thumb.” “[Consumers]

9 Examples include websites, AOL Instant MessengerSM

,, Kazaa Media Desktop (and other peer-to-peer sharing software), and many other non-commercial software packages. Packaging a product with advertising built-in and then offering a fully functional version without advertising (for normal sale) is a somewhat common practice for software manufacturers who do not produce full commercial products.

David Thaw University of Maryland, College Park Departmental Honors Thesis of 57 may know the seller’s or producer’s general reputation for expertise. In addition,

[consumers] may have reasons to presume the seller or producer exercises independent

judgment and to believe that this supplier users this independence to try to serve the

purchaser’s interests – reasons purportedly underwritten by professionalism in education,

law, psychology, or the priesthood” (2002 13).

These two examples of how individuals evaluate media and communications

producers illustrate a fundamental difference when applied to the computer software

market. With regard to the first “rule,” two problems arise. First, computer software is

such a young market that its businesses have not been in existence sufficiently long for

consumers to form meaningful opinions about their reputation. The tumultuous nature of

the information technology and computer software sector during the last several years

further exacerbates this problem. Second, in the area of computer software – particularly

computer operating systems – Microsoft Corporation exercises such a powerful

monopoly that individuals may find themselves compelled to form positive opinions

about it due to a mere lack of choices combined with Microsoft’s massive advertising

ability.

The second “rule” presents a very interesting set of concerns. The computer

software industry, unlike education, law, medicine, and engineering, is not currently

underwritten by established and universally accepted private-sector professionalism.

Equally troubling is the lack of state or federal laws requiring professionalism in

computer software development and manufacturing. Yet we continue to use computer

David Thaw University of Maryland, College Park Departmental Honors Thesis of 57 software in high-reliability systems,10 and it is increasingly becoming an integral part of

our communications infrastructure. Thus consumers have no means by which to evaluate

whether Microsoft Windows is an effective product for satisfying their needs, while at the

same time Microsoft Corporation is able to include in Windows “marketing” aspects (as

described in the above analysis of Baker’s third criterion) that may influence the user in

ways contrary to the ideals of equality and free exchange of information I examine as

essential to a healthy democratic society.

The application of Baker’s framework to modern information technologies,

specifically computer software, presents obvious concerns for democratic society as it is

evaluated in this paper. Pricing conditions, which will be examined later, could prevent

individuals from being able to express their preferences through the market. This

condition is further exacerbated by corporations’ (particularly Microsoft Corporation’s)

ability to derive revenue from third parties involved in the transaction by providing them

favorable positioning within their products.11 At the same time, however, the potential

positive externalities from choosing a software monopolist’s product further increase

pressure on consumers to utilize such products. Finally, the youth of software markets

and other factors discussed above give consumers little ability to objectively evaluate the

products they are purchasing.

10 High reliability systems tend to be subject to more strict internal guidelines for testing, system maintenance and stability that provide an increased level of security and operational guarantee not examined in this line of argument. 11 This phenomenon is not strictly limited to operating systems manufacturers. Computer system manufacturers such as Dell Computer Corporation engage in similar practices when they setup and ship their pre-installed systems. Microsoft has, however, engaged in what was described by many as “predatory” practices in attempting to force companies like Dell into selecting the applications Microsoft preferred. Furthermore, Microsoft has the “first crack” at embedding such “software/services advertising” into their operating system in a manner such that the computer system manufacturers cannot remove it.


The conclusion of this analysis is that an unregulated private-sector market

currently fails to provide consumers with the ability to express their preferences. Given

the potential information technologies have to influence the content of the information

they disseminate, therefore, and the inability of the market to guarantee the equality and

choice necessary for the principles of democratic society examined here, it is prudent to

examine how information technologies are developed to determine what alternative

organizational structures may be used.

Section V: Information Technology Development

In discussing the impact of modern information technologies on democracy,

therefore, it is important to examine the driving forces behind the development of these

technologies. For the purposes of this examination I will look at two categories of

producers: 1) private entities, in particular, non-governmental businesses; and 2) public

entities, in particular, the federal government. While businesses and the federal

government do not represent an exhaustive list of potential producers, in the case of

information technologies and information producers these categories encompass the vast

majority of producers.

In a capitalist society, such as the United States, it is reasonable to assume that a

private entity will not be motivated to incur significant cost to produce a good merely for

its social benefit. Private entities, rather, are generally motivated by the opportunity for

economic gain. For a business or corporation to have incentive to produce a good,

therefore, there must exist a potentially profitable market for the sale of that good. Goods

David Thaw University of Maryland, College Park Departmental Honors Thesis of 57 of this nature are most often excludable, and therefore comprise private goods and natural

monopolies.

Likewise, in a democratically governed capitalist society, it is reasonable to

assume that a public entity, in particular the federal government, will produce a good

when that entity is politically motivated to do so. In theory, this condition should occur

when the “will of the people” is such that a given good should be produced. For the

purposes of this discussion, an examination of whether this condition is actually the case

is not necessary. It is sufficient to recognize that a political motivation, such as the

election process or concerns of national welfare, will motivate the federal government to

produce a good. Goods of this nature therefore tend to be non-excludable, and are

produced in conditions where the market fails including public goods and common

resources. A relatively straightforward example is the creation of public highways

funded by state and the federal governments.

Given the importance of these technologies to democratic society, however, it

follows that their production and sale should be governed at least to some extent by the

“will of the people.” The protections and freedoms afforded private producers are a

product of the people granting those entities the freedoms of a capitalist government, and

as such it is reasonable for the citizenry to expect behavior becoming of participants in a

democratic society in return.12 In particular, a private producer should not be able to gain

monopolistic control of the production of a given information technology lest that

12 One example of this is the allocation of the frequency spectrum by the Federal Communications Commission. Specific slices of spectrum are allocated to different private entities by the FCC, giving these private entities a right to utilize them for things such as radio and TV broadcast. In return for this right, however, these private companies are expected to abide by certain regulatory practices and participate in such public activities as the maintenance and testing of the Emergency Broadcast System.

David Thaw University of Maryland, College Park Departmental Honors Thesis of 57 producer gain a potentially coercive power over society. Assuming that “government is

the sole institution with the monopoly rights on the use of coercion” (GVPT170H

01/1999), no private entity should be legitimately able to gain coercive power without the

consent of the citizenry. It therefore follows that in the case of natural monopolies, such

as the telephone system, public regulation is both a legitimate and necessary response to

protect the rights of the citizenry.

Given these various conditions, and the concerns raised in Section IV, it seems

prudent to examine the historical schema by which information technologies have been

produced. Assuming that we have achieved a reasonable degree of success thus far in

maintaining production organization schema conducive to democratic society, we can

then derive from these historical patterns information that can be applied to determine

what alternatives are available to address concerns with emerging modern information

technology systems production.

The Postal System

The postal system, strictly speaking, constituted a natural monopoly. Though

significant additional volumes of postal items could affect the system’s operation, one

individual’s use did not significantly reduce another’s ability to use the system. Use of

the system is an excludable good; a postal clerk could refuse to accept an individual’s

item or to deliver mail to an individual. Though natural monopolies tend to be produced

by private entities, in this case historical tradition resulted in the postal system being

instituted by the government. “In the seventeen-year period between 1775 and 1792, the

David Thaw University of Maryland, College Park Departmental Honors Thesis of 57 American postal system was little more than a mirror image of the royal postal system for

British North America as it had existed prior to 1775” (John 25). The rationale behind

this tradition lay in its fiscal success in Great Britain. “’There cannot be devised,’

[William] Blackstone observed, a ‘more eligible method… of raising money’” (John 26).

This tradition was maintained with the adoption of the federal Constitution, which

provided Congress with the power “[t]o establish Post Offices and post Roads.”

Eventually, as the necessity of guaranteeing “the citizenry the right to secure

uninterrupted access to up-to-date information about the ongoing affairs of state” (John

28) became clear to Congress, more attention was given to the postal system as a benefit

to democratic society. After much debate Congress passed the Post Office Act of 1792

that “broke radically and irrevocably with the inherited traditions of the past” (John 31)

and began the shift of the postal system from a revenue-generating source to an

instrument of American democracy. In addition to serving as a mechanism for citizens to

communicate directly with one another more easily, the federal government granted

special second-class mail privileges to newspapers. These privileges enabled newspapers

to be more widely distributed to individuals than would have been possible at the time by

providing a distribution infrastructure.13

13 Furthermore, this provided an infrastructure to individual who would not likely be connected if such infrastructure were to be developed privately. There would be little incentive to develop the network to distribute one newspaper a week to a backwoods farmer. With the postal system, however, that backwoods farmer could just receive this newspaper in the mail.

David Thaw University of Maryland, College Park Departmental Honors Thesis of 57 Broadcast Radio

Broadcast radio was another information technology challenge that democracies

faced. Broadcast radio provided a means by which information could be produced and

disseminated from a single location to an (effectively) unlimited number of consumers,14

provided they had the necessary equipment to receive it. “[By] 1922, more than 500

commercial broadcasting stations were licensed by Secretary of Commerce Herbert

Hoover” (Emord 143). In 1922, as a result of technological advances, a vast increase in

the number of radio broadcast stations was seen. “Erik Barnouw describes the prolific

growth in the industry:

So widespread was the feeling that broadcasting was a key to influence and power, that the rush was joined by many different interests. They included, not surprisingly, telephone and telegraph companies and the makers and sellers of electrical equipment. But educational institutions also took a prominent role. Scores of universities had long been active in radio experimentation and felt especially ready for this moment. By January 1923, broadcasting licenses had been obtained by 72 universities, colleges and schools, while many others prepared to follow. Newspapers were only very slightly less numerous; in the same period 69 newspapers became broadcasting licensees, along with 29 department stores, 12 religious organizations, several city governments, and a sprinkling of automobile dealers, theaters, and banks.” (Emord 144) This rapid and massive growth resembles the growth of the computing and

Internet industry in the late 1990s. A new technology had been introduced, and

organizations were becoming aware of its potential and rapidly moving to take advantage

of it. A crucial difference existed between the radio and computer industries, however –

the radio broadcast industry had already been and would continue to be subject to federal

regulation.

14 Within a given broadcast range.


In 1910, the United States Congress chose to exercise their regulatory power and

passed the Wireless Telegraphy Act. “[I]t was designed to improve maritime safety in

ship-to-shore and ship-to-ship communications by requiring all vessels that had more

than 50 people aboard and that traveled between ports that were more than 200 miles

apart to be equipped with efficient radio equipment” (Emord 140). This Act, however,

failed to address several concerns and did not provide for a nationalization of the radio

industry (140). Congress revisited the problem again in 1912, after the Titanic disaster,

passing a law that “prohibited the use of any apparatus for radio communication unless

the operator first obtained a license from the Secretary of Commerce and Labor” (141).

“This zero-priced system[, however,] quickly produced an excess demand for authority to

broadcast, causing [Secretary of Commerce] Hoover to withhold licenses on ground that

interference would result” (144). The U.S. Court of Appeals for the District of Columbia,

however, in Hoover v. Intercity Radio Company, ruled that the Secretary had exceeded

his statutory authority and lacked the power to withhold licenses. The result was a period

of four subsequent years during which most “regulation” of the industry was

accomplished by agreements among private entities. While the Secretary was able to

withhold licenses to some degree by restricting the available number of frequencies and

delaying action on license applications, federal regulation during this period (1923-1926)

was extremely limited. (Emord 140-146)

Congress reasserted itself in 1926. “In March 1926, the House Committee on the

Merchant Marine and Fisheries began consideration of a bill for comprehensive federal

regulation of radio” (Emord 167). Congressman Wallace H. White sponsored this bill.

White argued that “[d]espite statements made by Department of Commerce Solicitor

David Thaw University of Maryland, College Park Departmental Honors Thesis of 57 Stephen B. Davis, Jr., that broadcast license were held by a diverse number of interests…

an alleged monopoly [was] arising in the manufacture of radio receiving sets and [was]

on the verge of taking over the entire broadcasting industry” (167). White’s arguments

were supported by Secretary Hoover, who “asked the committee to treat the businesses

engaged in radio broadcasting as utilities that must not be conducted for private gain but

must be forced to serve the ‘public interest’” (167). Other Members of Congress

expressed similar concerns for this “public interest.” “Congressman William R. Johnson

predicted:

If the strong arm of the law does not prevent monopoly ownership and make discrimination by such stations illegal, American thought and American politics will be largely at the mercy of those who operate these stations. For publicity is the most powerful weapon that can be wielded in a Republic, and when such a weapon is placed in the hands of one, or a single selfish group is permitted to either tacitly or otherwise acquire ownership and dominate those who dare to differ with them, it will be impossible to compete with them in the ears of the American people.” (Emord 170)

After extensive debate on both sides of the issue, Congress passed a compromise entitled

the Radio Act of 1927 that vested licensing authority in the hands of a newly created

Federal Radio Commission. This authority, while only temporary, was renewed by

Congress until 1934 when the Communications Act of 1934 created the Federal

Communications Commission. (Emord 175-183)

Broadcast Radio As Compared To Computer Information Technologies

The example of the radio broadcasting industry, the role of government in its

regulation, and the debates surrounding this regulation provide some of the most striking

historical parallels to the current information technology revolution. Similar to the events

David Thaw University of Maryland, College Park Departmental Honors Thesis of 57 of recent years, a relatively nascent technology began to explode in popularity as private

entities realized its potential. Radio broadcast technology had been present prior to 1922;

however it was relatively underutilized by the private sector for commercial purposes.

Likewise, computing technologies had been present long before the information

technology revolution of the late 1990s, but were relatively underutilized by the private

sector. In each of these cases, a technological “boom” then ensued in which myriad

organizations began to acquire interest in and utilize the technology. A crucial difference

between these two examples, however, exists. While in both cases the technology was

produced primarily by private corporations, in the case of the computer revolution

virtually no regulation existed (or currently exists).

It is therefore important to determine why this difference exists and what

implications the difference has for democratic society. I assert that the causal factors of

the difference do not bear relevance to this argument.15 Rather, it is important to focus on

the implications of the difference. Congressman White described a monopoly in the

manufacture of broadcast technology. I would assert that this technology, as will be seen

later in Section VI, is constructively similar to the computer operating systems

technologies in use today. Accordingly, we can then infer that many of the same

15 The use of radio in Naval affairs and operations, combined with the extreme difficulty in selectively restricting radio use by force, led the Navy and associated maritime organizations to pressure Congress to regulate the airwaves. Since these organizations were among the forerunners in the use of radio technologies, and these requests were made nearly a decade prior to the commercial radio revolution, a precedent had already been established for federal regulation in this area (Emord 137-141). While these factors may explain the difference in regulatory involvement between the radio and computer revolutions, they do not provide meaningful insight into the implications of the lack of regulation for democratic society.

David Thaw University of Maryland, College Park Departmental Honors Thesis of 57 arguments expressed in justifying the regulation of the radio industry would apply to the

computer industry should similar monopolistic conditions arise.

Modern Information Technologies

The microprocessor-based information technology industry, particularly the PC

sector, is economically unique in that it comprises primarily unregulated private-sector

producers. Furthermore, unlike broadcast television and radio, many of the goods

produced by this industry are excludable, resulting in a shift in the industry from public

goods production back to natural monopoly production. Early American information

technologies, as discussed above, comprised primarily the printing press and the postal

system. Both technologies are effectively non-rival, as one person’s use of the postal

system does not significantly reduce another’s ability to effectively use the system.

Likewise, one person’s reading the material in a printed copy of a newspaper does not

effectively reduce another person’s ability to do so.16 Unlike broadcast media, however,

both of these “goods” were excludable. A postal agent could refuse to deliver an

individual’s mail, and a newspaper could refuse to sell an individual a copy of its

publication. Broadcast media, in contrast, could be viewed or listened to by any

individual with the appropriate hardware to receive the broadcast.

Information technologies have advanced significantly over the course of the past

decade due largely to advances in computing technologies. Personal Computers (PCs)

and other microprocessor-based devices have rapidly consumed portions of the market

16 In fact, more readers could actually enhance the product for all by reducing the per-unit cost of production and distribution.

David Thaw University of Maryland, College Park Departmental Honors Thesis of 57 for information distribution technologies. While such devices have been more widely

adopted in some information content markets than in others, nearly all information

markets have seen a significant increase in the use of computing technologies over the

past decade. It is therefore prudent to examine the impact these emerging technologies

may have on personal liberty and democratic society given the importance of information

and its distribution to these concepts of democracy.

Conclusions – Criteria For Impacting Information Flow and Democratic Society

Based on the analysis above, for an information technology to be able to impact

personal liberty and/or democratic processes three conditions must be met. The

technology must enable those who control it to influence either the content or availability

of the information it distributes; those who control it must have incentive to influence the

content/availability of that information;17 and there must not exist sufficient alternatives

for consumers such that they could seek the same information provided by that

technology elsewhere.

Section VI: Understanding the Operating System

Influencing Information Flow

To examine if a technology enables those who control it to influence the content

or availability of the information it distributes, we must first characterize the manner in

which the technology functions with respect to information. In the case of modern

17 This incentive may/may not be direct. It may be indirect, as in the case of product placement in TV/Movies.

David Thaw University of Maryland, College Park Departmental Honors Thesis of 57 information technologies, we focus on the computer. Computing technologies can be

broken down into layers of abstraction. Silberschatz, Galvin, and Gagne argue that

computing systems can be divided into four basic layers. These layers form a

technological hierarchy, through which the computing system operates and produces

results. The four components are: the hardware, the operating system, the application

programs, and the users (Silberschatz 3).

Understanding how these layers interact is crucial to determining if a producer of

a given technology can exert influence over the content or availability of the information

that technology distributes. In most modern computing systems, the three technological

layers – the hardware, the operating system, and the application programs – are not

produced by any single entity.18 The components that comprise these layers, rather, tend

to be manufactured by a variety of different producers. At the hardware layer, for

example, different manufacturers produce the central processing unit (CPU), the main

memory, the storage devices, and the video processing and rendering hardware for most

PCs. Likewise, at the application layer, most modern computers are sold with software

from at least two different manufacturers (many are sold with more). The Operating

System layer, however, is different in that its components generally comprise a single

inseparable package that is at least sold, if not also produced, by a single entity.

Furthermore, there are relatively few commercial producers of operating systems

currently in business.

18 Certain efforts by Microsoft Corporation are starting to defy this model. For the purposes of discussion within this section, I do not address the efforts.

David Thaw University of Maryland, College Park Departmental Honors Thesis of 57 Defining The Operating System – Establishing A Working Definition

In order to understand how Operating Systems may be used to influence the flow

of information to the end user of a computing system, we must establish a working

definition of what the operating system encompasses. There are several different

definitions of what comprises an operating system. One viewpoint examines an

operating system as a “resource allocator.” “A computer system has many resources

(hardware and software [or application]) that may be required to solve a problem: CPU

time, memory space, file-storage space, I/O devices…” (Silberschatz 4). Each of the

resources listed here comprises an element of the hardware layer. In this viewpoint,

“[t]he operating systems acts as the manager of these resources and allocates them to

specific programs and users as necessary for tasks” (Silberschatz 4). In essence, this

view looks at the operating system in terms of its management of the hardware, and does

not specifically address the distinctions between the applications and the operating

system. This approach is important in that it does not limit the operating system to a

specific set of functions, and fails to draw a clear line between the operating system layer

and the application layer. Given the prospect of potential monopoly in operating systems

technology as affording OS producers the ability to influence the availability and/or

content of the information their systems provide, an expanding definition of the operating

system permitted by this viewpoint presents a significant concern.

A more restrictive definition of an operating system, in contrast, describes it as

“the one program running at all times on the computer (usually called the kernel)”

(Silberschatz 5). This is a much more common definition in technical literature, though it

may fail to reflect the average user’s perception of an operating system. Furthermore,

David Thaw University of Maryland, College Park Departmental Honors Thesis of 57 given that the nature of the products actually marketed and sold by operating systems

manufacturers almost always comprise far more than just the kernel, this restrictive

definition is likely insufficient to use in examining the potential for a producer to exert

influence over the information its system distributes.

For the purposes of this argument, a combination of the definitions presented in

the literature will be used. This approach is primarily due to the practical implementation

of operating system products available at the time of writing. We consider three primary

manufacturers of computer operating systems: Microsoft Corporation, Apple Computer,

and Sun Microsystems. Each of these manufacturers possesses a non-trivial market

share, and was selected to represent a class of similar operating systems. In the case of

each manufacturer, their operating system product includes three main sets of

components: 1) the core operating system components, 2) a suite of utilities, and 3) an

interface shell (usually comprising a Graphical User Interface or GUI). I propose these

divisions as a way of modeling the shape of operating systems actually in use and on the

market today. These divisions are not intended to be restrictive; in fact, there are several

areas of overlap among them. Rather, they are designed to help define and classify what

constitutes an “operating system” in modern PC markets.

The Core Operating System Elements

Describing the elements of the core operating system components is a comparatively

easy task. Silberschatz breaks these down into eight pieces (Silberschatz 46-51):

• Process Management • Main-Memory Management


• File Management • I/O System Management • Secondary Storage Management • Networking/Communications • Protection Systems/Security • Command-Interpreter System(s)

Understanding the elements of the core operating system components is crucial because,

as the foundation elements of an operating system, these will be the first to be

implemented as operating systems are developed for expanding markets such as PDAs,

mobile phones, and gaming systems. In particular, the Process Management,

Networking/Communications, and Protection Systems/Security elements are of particular

concern to the issue of information flow in computing systems. The potential manner in

which these elements can affect information flow to the end-user is discussed further in

Section VII.

More restrictive definitions of operating systems, as discussed above, are

generally limited to components in this category. Prior to the 1990s, these definitions

may have been more effective in capturing the nature of the products produced by OS

manufacturers. MS-DOS versions 3 through 6, for example, qualified these more

restrictive definitions of operating systems while providing limited additional

functionality. Additional functionality (particularly in MS-DOS versions 5 and earlier)

such as graphical user interfaces, third-party hardware device and peripheral drivers, and

network communication required the installation of additional software. Often times this

software was not available from the OS vendor. A similar case can be made for UNIX-

variant operating systems; however it is important to note that Apple Computer’s

operating systems began to add some of these additional features in the early 1980s with

David Thaw University of Maryland, College Park Departmental Honors Thesis of 57 the introduction of the Apple Macintosh computer. (Silberschatz 6-11, 19-21)19 From a

practical standpoint it is clear – and is consistent with definitions in the literature

(Silberschatz 46-51) – that these factors must be included in a working definition of an

operating system.

The Suite of Utilities

The suite of utilities provided with an operating system varies by manufacturer,

but historically such suites have included advanced storage management utilities, local

area and internet-capable networking protocols, and drivers/controllers for standard

peripherals. Examples include hard-disk defragmentation utilities, built-in TCP/IP

protocol stacks and associated modem control software, and standard printer, hard disk,

floppy disk, display, keyboard, and mouse drivers. While the manufacturer sometimes

provided these drivers for inclusion with the operating system, they were almost always

distributed with the operating system. In the case of Microsoft operating systems, the

manufacturer-provided approach was most common. Apple Computer’s systems were

far more tightly integrated than “Intel-compatible” systems (as defined by Judge

Jackson’s ruling) and therefore Apple developed many of these utilities themselves (84 F.

Supp. 2d 9).

Among the more crucial of these utilities is a set of software applications I will

define as “information and media access software.” These include applications for

19 Much of the information provided in this section comes from knowledge I have acquired working with computing systems over the course of the last ten years. While some support for it is provided in the literature, the primary source of information is my own hands-on experience working with these systems.

David Thaw University of Maryland, College Park Departmental Honors Thesis of 57 accessing the World Wide Web, viewing multimedia content, and accessing basic (non-

web-based) email services. While these applications clearly do not fall under the strict

definitions of an operating system, Microsoft Corporation has been including them in

their operating systems packages since the mid-1990s. These applications represent one

of the core potentials of how an operating system can influence the flow of information.

If such applications – which provide the essential means by which end-users access most

types of information online – are considered part of the operating system, then the

manufacturer of the operating system would effectively control which such applications

ran on a given user’s machine. Indeed, it is the case, as shown by the findings of fact in

Judge Jackson’s ruling, that Microsoft operating systems include web-browsing

applications as part of their standard package (84 F. Supp. 2d 9). Furthermore, “[i]n

contrast to other operating systems vendors, Microsoft both refused to license its

operating system without a browser and imposed restrictions – at first contractual and

later technical – on OEMs’ [Other Electronics Manufacturers] and end users’ ability to

remove its browser from its operating system” (84 F. Supp. 2d 9).

Furthermore, Microsoft decided in 1997 to aggressively pursue the integration of

multimedia/streaming media applications into their operating system (84 F. Supp. 2d 9).

While pursuit of this integration has not been as strong amongst Microsoft’s competitors,

it has occurred to some degree with Apple Computer’s “QuickTime” multimedia access

application (Apple – QuickTime Player). Given the nature of these factors, it is clear that

a practical working definition of an operating system must include the suite of utilities

provided by major operating systems manufacturers. Furthermore, given Microsoft’s

massive market share, I assert that information and media access software must be

David Thaw University of Maryland, College Park Departmental Honors Thesis of 57 included in this suite of utilities and therefore considered part of the operating system as

well.20

The Interface Shell (GUI)

The Interface Shell is an example of where some of the distinctions between the

operating system component classifications I propose blur. For any computing system to

function, it must be able to receive and process commands from the user. Returning to

the hierarchical model of a computing system discussed above, we can now propose a

fourth layer that sits above the applications layer – the “End User” (Silberschatz 3-4).

Following from the condition then that layers in this hierarchical model generally interact

only with those that are directly adjacent to them, we can conclude that there must exist

an application program through which the user can interact with the operating system.

One of the core elements of the operating system discussed by Silberschatz is the

Command-Interpreter System. “[It is] one of the most important systems programs for an

operating system, [acting] as the interface between the user and the operating system”

(Silberschatz 51). In its most basic form, the command interpreter consists of a text-

based input (often referred to as a “command-line”) into which the user can type

commands that the operating system will execute. While “[s]ome operating systems

include the command interpreter in the kernel, [o]ther operating systems, such as MS-

DOS and UNIX, treat the command interpreter as a special program that is running when

20 It should be noted that my assertion is for the purposes of a working definition, to be used in the analysis of how an operating systems monopolist might affect the flow of information to and end-user of a PC. This statement is not intended to make any assertion as to the proper or technically correct state of whether such utilities and applications should be included in the definition of an operating system.

David Thaw University of Maryland, College Park Departmental Honors Thesis of 57 a job is initiated, or when a user first logs on [or the computer is first activated]”

(Silberschatz 51).

This logic, on the surface, seems counterintuitive to my point that the Interface

Shell should be included in a working definition of an operating system. In fact,

however, it represents an argument for why this working definition must be expanded to

address the practicalities of modern OS implementations. Particularly in the case of more

recent Microsoft operating systems (Windows 95 and later), it is virtually impossible for

the user to run or use any software applications if the interface shell is not functioning.21

While the Interface Shell, therefore, is technically an application program, it is one

without which the operating system rendered effectively useless to the end-user. It

therefore must be included in this working definition of the operating system.

The Resultant “Working Definition”

It can be concluded from this analysis that elements of the three main components

included in each of the primary OS versions discussed above must be included in a

working definition of an operating system.22 For the purposes of this argument, then, an

operating system will therefore be considered to be that portion of software, generally

sold/delivered as a single package, produced by a single manufacturer, that is required for

any application program to run on a PC. It will include all aspects of that software 21 A relatively easy-to-accomplish demonstration of this phenomenon can be performed on nearly any Windows NT, 2000, or XP machine by manually closing the user shell (“explorer.exe”) process via the Task Manager application. This procedure, however, is not recommended except to be performed by experienced users with the knowledge to manually restart the explorer interface shell. 22 This working definition is intended for computer operating systems, and may not be extensive to expanding operating systems markets such as PDAs, mobile phones, and gaming systems. I would assert that the use of (non-proprietary) “software-based” operating systems in such devices is too new a phenomenon for there to be sufficient data from which to draw meaningful conclusions.

David Thaw University of Maryland, College Park Departmental Honors Thesis of 57 relevant to the control of hardware, application management, security, and the

accompanying (usually graphical) user interface(s) to access these functions.

Furthermore, it will include any information and media access software included with the

operating system unless that software is clearly distinct and can be easily removed by the

average end-user. Adopting then a broader definition of an operating system, we can

examine how manufacturers could use their technology (whether intentionally or not) to

influence information flow.

Section VII: How Computer Operating Systems Can Influence Information Flow

In its simplest form, a computer is a device for storing, manipulating, processing,

distributing, and accessing information. It is a true “information technology” in this

sense. While computing devices have been adopted to perform a wide variety of

functions, we focus on the storage, processing, and transmission of information in this

analysis. Extrapolating then from the working definition of an operating system

established above, we find several ways in which it can influence information flow in a

computing system.

Direct Influences

Direct influences result from situations when the operating system directs or

forces the user to utilize a specific information source. Microsoft Windows XP, for

example, provides a very easy-to-use interface for establishing an Internet connection

when a user first sets up his/her computer. One of the steps in this “Wizard” (see Figure

I) provides the user with the option to choose from a selection of means of connecting to

David Thaw University of Maryland, College Park Departmental Honors Thesis of 57 the Internet. One of the selections on this list is “Choose from a list of Internet Service

Providers (ISPs).” This is the top item of the list, and also the default option. If the user

selects this option, he/she is then taken to another screen with two primary options: 1)

“Get online with MSN”, and 2) “Select from a list of other ISPs.”23 MSN (loosely)

stands for “Microsoft Network” and is Microsoft Corporation’s primary Internet Service

Provider. It is a direct competitor with America Online (AOL), and a user’s selection of

either service provider will significantly affect how he/she accesses information through

the Internet, as each of these services have their own user interfaces and software for

aggregating, analyzing, searching, and retrieving information from the Internet. Thus one

direct method by which an operating system can influence information flow is by leading

the user to select a specific means for establishing communications connectivity.

Given that the operating system includes information and media access software, and

that the operating system controls which applications are used to perform different data

access functions, it therefore can effectively direct a first-time user to use specific

applications to retrieve information from the Internet and to access information that user

possesses.24 Furthermore, since the operating system controls the execution of programs

and since the manufacturer of the operating system does not have to release all of its

source code, the manufacturer could advantage certain programs over others in terms of

efficient and stable execution. Microsoft has been repeatedly accused of doing this with

their Internet Explorer web browser to the detriment of Netscape Navigator, a competing

23 In the default setting for Windows XP Professional, the “other ISPs” list is empty. This may vary depending upon the computer manufacturer and what other ISPs they add to the configuration (if any). It should be noted, however, that Microsoft could add any additional ISPs they chose into the software and that MSN is always included. 24 Information possessed by the user on other media such as CDs, DVDs, floppy and zip drives, etc.

David Thaw University of Maryland, College Park Departmental Honors Thesis of 57 web browser (84 F. Supp. 2d 9). The choice of an information and media access

application is not insignificant to a user’s experience. Certain multimedia playback

software, for example, cannot open certain video files. Microsoft’s Windows Media

Player, for example, does not support “MOV” (QuickTime) format video files.25 In

another example, Microsoft Office XP’s “Outlook” application, which is (amongst other

things) the email client in its office suite, blocks certain types of file attachments as part

of its default implementation. When operating as a standalone client, Outlook XP does

not allow this setting to be changed.26 Outlook provides this “functionality” in order to

curb the spread of computer viruses, by blocking specific file types commonly known to

transmit viruses. In doing so, however, Outlook is literally censoring the information a

user receives. While Outlook is not a part of the Windows operating system, Windows

does include a more “stripped-down” version of it, known as “Outlook Express.” As of

version 6, Outlook Express did not include this functionality. If Microsoft were to

include this functionality in later versions of Outlook Express, however, it would

effectively be using its operating systems monopoly to censor a potential user’s email.

Thus clear potential exists for an operating system monopolist to use information and

media access applications to influence the flow of information through a given computer.

The Graphical User Interface (GUI) Shell is another means through which the

operating system can influence the flow of information in a computer. The GUI Shell is

actually what the user sees and interacts with when using the computer. It was first seen 25 This is effective as of Windows Media Player Versions 8 and earlier. This refers to the default configuration – it may be possible to have Windows Media Player support these files; I have not been able to discover how. 26 I have heard – informally – that there may be a way to override this. After extensive research in Microsoft’s technical support and knowledge bases in August 2002, I was unable to determine any method of doing so.

David Thaw University of Maryland, College Park Departmental Honors Thesis of 57 prominently on the Apple Macintosh computer in the early 1980s, and was designed to

replace the Command Interpreter as the sole method of allowing a user to operate the

computer and execute commands. Since then, the GUI Shell has developed into an

extremely complex program that presents and organizes information to the user about the

computer’s operation and its contents. In essence, the GUI informs the user about how to

operate the computer – an information storage and processing device – and provides

(essentially) the sole means through which this can be accomplished. By manipulating

the operating system’s GUI, therefore, an OS manufacturer can exert influence over the

experience the user has with the computer and the manner in which he/she accesses,

stores, and manipulates information.

Indirect Influences

Equally important, and perhaps more interesting are the indirect ways in which

operating systems can influence information flow. There are many types of ways in

which such influence can occur; for the purposes of this examination I will focus on those

derived from the concept of the operating system as the “governor” of a computing

system.

As discussed above in the “layered model” of a computing system, the operating

system isolates the user from the hardware of a computing system.27 The user does not

directly manipulate the hardware, nor does the user (in general) run application software

(other than the operating system) that does so. Rather, the user runs applications

27 This refers to interface-based control, not simple operations such as opening or closing the CD tray or moving the mouse. However, it should be noted, that in newer operating systems even these operations are somewhat isolated from the user through the operating system.

David Thaw University of Maryland, College Park Departmental Honors Thesis of 57 designed to function on a specific operating system. These applications require the

resources of the computer hardware and, as discussed above, access it solely through the

operating system. This is generally accomplished through an Application Programming

Interface (API). An API specifies what functions, or “tools,” are available to a

programmer designing a software application and how to implement or “use” these tools.

Loosely speaking, each operating system has its own API. In general, modern software

applications are developed for a given platform (operating system) and use the resources

of that platform solely through the API.

The API clearly plays an important role in the development of computer software.

It is clear that an operating systems manufacturer could use the API to affect what a

computer was able to do, but the question of how and why the manufacturer would

accomplish this remains. The answer to this question lies in the lack of necessity for

operating systems manufacturers to expose (or make publicly available) their entire APIs.

If an OS manufacturer were only to make publicly available a portion of its API, then

other software manufacturers who wished to develop applications for that manufacturer’s

OS platform would be at a severe disadvantage against those to whom the manufacturer

had exposed the entire (or larger part of) API.

The means discussed here are hardly an exhaustive list of how operating systems

manufacturers can use their products to influence information flow in a computing

system. The important point to note is that there exists great potential for doing so.

While it is unclear whether any of the concrete examples discussed in this section are

actually affecting information flow to end-users in a manner significant to democracy, the

David Thaw University of Maryland, College Park Departmental Honors Thesis of 57 clear potential for this to occur suggests that the question of whether manufacturers have

incentive to engage in this type of behavior should be examined.

Section VIII: Incentives To Influence

The incentives for operating systems manufacturers to influence the flow of

information derive from a combination of the current organization of production (i.e.,

unregulated markets) and my extension of Baker’s theories on media and

communications markets. For the purposes of this discussion, I will exclude the more

grim possibilities of ideologically motivated corporations that seek to influence

individuals for personal as opposed to economic reasons.

Baker’s third criterion, as discussed in Section IV, describes a trinary sales

relationship in media products. I have asserted that a similar relationship exists in the

manufacture of computer software, especially that of computer operating systems. In an

unregulated market, a corporation’s primary objective will be to maximize profit. While

still a member of a society, that corporation (provided it maintains legal activity) is held

responsible only to its shareholders, who presumably will desire increased profits. The

operating system manufacturer’s primary goal, therefore, is to generate revenue as

efficiently as possible.

There are two potential cases for an operating systems manufacturer: 1) an OS

manufacturer whose primary product is their operating system and who does not have a

meaningful investment in other software markets; and 2) an OS manufacturer who does

have a meaningful investment in other software markets. While other permutations of

investment and product breadth are of course possible, this categorization will better

David Thaw University of Maryland, College Park Departmental Honors Thesis of 57 characterize the potential incentives a manufacturer could have to influence information

flow. The first case, I would assert, is moot at this point as there are no significant

examples of such a company. It would be worth examining, perhaps, had Judge

Jackson’s proposed remedy of dividing Microsoft Corporation been upheld, however this

was not the case, and the settlement approved by the Justice Department does not include

such a remedy. Rather, then, we turn to the second case, under which Microsoft

Corporation, Apple Computer, and Sun Microsystems would fall.

Given their overwhelming monopoly, Microsoft is clearly the current example

most worth examining.28 In addition to operating systems (which as discussed earlier we

can define to include several information and media access applications), Microsoft also

produces several major software applications packages, most notably its “Microsoft

Office” package of standard business applications. Microsoft therefore has a vested

interest in promoting these products. Furthermore, as an operating systems manufacturer,

Microsoft has a vested interest in promoting the introduction of its operating system into

emerging computing markets such as those for Personal Digital Assistants (PDAs),

gaming systems, mobile telephones, and other devices that may at some time incorporate

a software-based operating system. Microsoft therefore has strong incentive to use their

operating systems monopoly to influence the flow of information in a way that promotes

these business ventures. The above example of how Microsoft promotes the use of its

MSN Internet service, is one instance of how the company is engaging in such practices

28 A comparative analysis of other operating systems manufacturers might prove worthwhile, however such an analysis is outside the scope of this paper. It should be noted, however, that if market trends shift over the course of the next several years and the market becomes less monopolistic, such a comparative analysis would likely become important.

David Thaw University of Maryland, College Park Departmental Honors Thesis of 57 today. An extensive list of specific business practices Microsoft engaged in, both

through the use of their operating system technology and through its software monopoly,

is examined in extensive detail in Judge Jackson’s Findings of Fact (84 F. Supp. 2d 9).

Microsoft’s actions included leveraging its monopoly against firms such as IBM, Apple

Computer, and Real Networks.

Section IX: Insufficient Alternatives – The Shape Of The Market Today

Given the potential for operating systems to influence the flow of information in

computing systems, the importance of computers to the exchange of information in

modern society, the incentives for OS manufacturers to use this potential to influence the

flow of information, and the importance of this information exchange to democratic

society, it is clearly important to examine the current state of the market to attempt to

characterize Microsoft Corporation’s monopoly and determine whether it presents any

real threats to democratic society. The original goal of this paper was to produce exactly

that characterization and, based upon its results, attempt to design a series of policy

recommendations that would protect the principles of democratic society discussed here.

In the course of conducting my research, however, I discovered that this question is

somewhat premature.

First, computer software markets are extremely young. Microsoft Corporation is

less than 30 years old, and computer systems role in individuals’ information exchange

did not begin to reach substantial percentage until the 1990s. This presents an immediate

problem in that with such young markets, it is unlikely that sufficient market data will be

available to predict long-term outcomes. Second, collecting data on computer software

David Thaw University of Maryland, College Park Departmental Honors Thesis of 57 usage at-large is an extremely difficult task. In the course of my research, I examined

resources including numerous journals, electronic databases, and texts.29 I was unable to

find reliable, comprehensive data regarding the utilization of various computer software.

The Problem of Analyzing Operating Systems Markets

Operating Systems Markets are particularly difficult to analyze for several

reasons. Most of these reasons derive from the earlier discussed point of the negligible

marginal cost of production of computer software. This problem is further exacerbated in

environments where end-users do not need the actual physical media (usually CD/DVD-

ROM) on which the software is distributed or their own individual copy of the packaging

reference materials, as is often the case with large organizations.

First, continually changing licensing structures make point-of-sale transactions

difficult to calculate. Site licenses, licenses that allow an organization to freely copy and

install a specific (or unlimited) number of copies of software on machines used by

members of that organization for purposes specified within the license, can make

calculation of the exact number of computers using a given software package difficult.

While in some cases these licenses are sold with a specific limit, it is difficult to

determine how this number compares to the actual number of copies in use. Point-of-sale

based tracking also may suffer from the question of whether a user is actually using the

copy of a given software package he/she purchased, how long he/she continues to use it

for, whether he/she uses more than one software package and the ratio with which he/she

29 Sources examined included, but were not limited to: Gartner Research, comScore/MediaMetrix (non-commercial), Gale Group’s “Business and Company Resource Center,” Hoover’s Online, LexisNexis, ProQuest “MarketResearch.com Academic,” and multiple RDS Business and Industry databases.

David Thaw University of Maryland, College Park Departmental Honors Thesis of 57 uses those packages. These factors matter because an unused software package

(presumably) will not be able to affect the flow of information in a given user’s

computing system.

Third, software piracy, a practice frequently decried by computer and software

manufacturers, has real impact on the number of copies of software in use. Illegally

copied software is virtually impossible to track, as users have no incentive to provide

information about their activities in this regard. Although unable to prove exact numbers,

estimates have ranged in the billions of dollars lost to the software industry per year over

the past few years due to software piracy.30

Fourth, “Open Source” software presents an almost impossible-to-track problem.

Open-source OS usage is difficult to reliably track due to the free availability of the

software (i.e., there is no “point-of-purchase” transaction to use in collecting market

data). While it might be possible to create sample populations and conduct surveys, I

was unable to find sufficient information in the course of my research to suggest that

there is sufficient knowledge to develop effective survey methodology.31 The

development of such methodology, if proven statistically valid, would present a solution

to many of the problems above.

30 See works Cited and Consulted: “Seventh Annual BSA Global Software Piracy Study,” “Software Piracy Costs Businesses Billions,” “Software Piracy Costs Billions.” 31 It may be the case that such methodology exists. I did not conduct as extensive a search into this as other areas of my research, however no obvious indicators of its existence were found.


An Introductory Analysis of Operating Systems Markets

Given that sufficient data are not available to analyze operating systems markets

in a manner sufficient for the original goals of this paper, I have alternatively conducted

an introductory analysis of the effects of current Microsoft pricing structures on the

individual consumer. This analysis is neither intended to be comprehensive nor

statistically sound, rather it is intended to introduce the reader to the problem of

individuals begin unable to express their “preferences” (as described in Section IV) due

to monopolistic conditions. It examines one of Baker’s proposed solutions to the

problem of marginal versus average cost pricing in media products, and discusses how

Microsoft fits into this model.

With the current state of socioeconomic affairs, it is likely that individuals will

experience a great deal of inequality in their ability to express their “preferences” through

the market. Microsoft Corporation, for example, clearly does not suffer from a condition

of underproduction – while I was unable to collect data to confirm it, it seems unlikely

that they are charging their customers the marginal cost. While they have implemented

some degree of price discrimination (i.e., with different versions of Windows such as XP

Professional and XP Home), the overall price), this discrimination is unlikely to

significantly differentiate among classes of consumers. The price difference between

Windows XP Professional ($300 US) and Windows XP Home ($200 US) represents a

33% decrease. While this differentiation does represent a savings to the consumer, I

would assert that an individual who is unable/unwilling to spend $300 for a necessary

product will not have their ability to do so affected by reducing the price to $200,

David Thaw University of Maryland, College Park Departmental Honors Thesis of 57 particularly in modern American culture where lines of credit (particularly through credit

cards) are easily obtained. A similar argument can be applied to the price discrimination

between the upgrade and full versions, which at best represents a 50% reduction in cost

(for Windows XP Home version) and requires a previous purchase of an operating

system at full price. (“Windows XP: How To Buy”). The result is a situation in which

consumers are beholden to pay the price that Microsoft determines.

Section X: Conclusions

This is a paper about the need to conduct research. It attempted to identify a

potential problem that may affect democratic society, and examines under what

conditions that problem could manifest itself. It then examined the current conditions,

and based on those conditions makes suggestions about what types of research should be

conducted to see if those conditions will manifest as meaningful threats to democracy in

the future.

While plagued by a lack of sufficient valid data on the specifics of the current

market situation, my research and experience leads me to conclude two important points.

First, there is a real, demonstrable potential for an operating systems monopolist to exert

influence over the flow of information through society. Computer operating systems

form the foundation of computing systems, which are becoming increasingly integrated

into our society. As effective “governors” of these technologies, operating systems

become much like the radio broadcast equipment assailed by Congressman White in the

1920s. They are powerful tools that enable collaboration, innovation, and can bring great

David Thaw University of Maryland, College Park Departmental Honors Thesis of 57 benefit to society and the economy, yet at the same time bring important responsibility to

those who produce and control them.

Second, Microsoft Corporation, while possessing a very powerful monopoly in

the operating systems market, does not pose a real threat to democratic society today. I

again return to my earlier point that this is a very young and rapidly changing industry,

one about which we have too little data to predict the future with great certainty. I

believe, however, based on my research and experience that it is the case that Microsoft

could develop into a potential threat if their market share holds and they remain

unregulated. The key elements exist – ability, motive, and a history of actions indicative

of their willingness to act on these abilities and motives.32

The Need For Interdisciplinary Research

As discussed earlier, this is a paper about identifying a problem and asserting the

need to conduct research about it. In particular, this paper is about the need for political

and other social scientists to be concerned with developments in the computer and

information technology industry. Their need for concern arises from the issues discussed

in this paper, in particular, for political scientists, the issues related to effective

democratic governance and maintenance of a democratic society.

This research, however, cannot be conducted in isolation. The efforts involved in

writing this paper, for example, involved the support of computer scientists as well as

political scientists. My assertion is not that there is some type of discipline-changing

32 See Judge Jackson’s opinions and findings of fact (84 F. Supp. 2d 9). Specifically, Judge Jackson outlines Microsoft’s actions toward companies such as Apple Computer, IBM, Netscape Communications, AOL, and Real Networks. Conclusions of Law and Order may also be useful.

David Thaw University of Maryland, College Park Departmental Honors Thesis of 57 revolution occurring, rather, it is that as new information technologies permeate our lives

with greater breadth and depth that social and technical scientists must work together to

characterize, study, and understand the impacts of these technologies. It matters to the

social scientist for her concerns about how technologies will impact questions that she

studies. It matters to the technical scientist for her concerns about how her research will

be received by, implemented in, and directly affects society.

Potential Research Avenues

Throughout the course of this paper I have suggested several potential research

avenues that resulted from unanswered questions encountered in my research. Most of

these questions focus on the lack of data available about the nature and shape of

computing and information technology markets, especially for software. In addition, I

have included several other “Unanswered Questions” that came up during the period I

was conducting my research.

It is my hope that this work will inspire others to take seriously the types of

questions I have raised and consider the other types of associated work that may be

conducted.

Unanswered Questions

Microsoft Corporation is currently pushing a strong initiative for its “.NET”

framework of Information Technology solutions. “Microsoft® .NET is a set of Microsoft

software technologies for connecting information, people, systems, and devices. It

enables a high level of software integration through the use of XML Web services—

David Thaw University of Maryland, College Park Departmental Honors Thesis of 57 small, discrete, building-block applications that connect to each other as well as to other,

larger applications over the Internet” (“Defining the Basic Elements of .NET”).

Microsoft has expanded the reach of this effort into most of its prominent software and

services products, including Microsoft Windows XP, the popular MSN Hotmail email

service, and several of its web server operating systems and their associated development

platforms. Clearly encompassing a massive portion of Microsoft’s efforts in the field of

information technology, .NET represents a potentially important avenue of research for

social and technical sciences concerned with issued related to the exchange of

information in global society.

In February 2003 WalMart Stores, Inc. announced that it would begin selling

retail copies of a Linux-based operating system as an alternative to Microsoft Windows

(“WalMart.com to Sell Lindows OS”). “LindowsOS is a new, fun and exciting operating

system that delivers the power, stability and cost-savings of Linux with the ease of a

windows environment. LindowsOS is easy to experience because of its friendly graphical

interface and support for popular Microsoft Windows file types (.doc, .xls, .ppt, .jpg, .gif,

.mp3, etc.)” (“LindowsOS”). Additionally, WalMart announced that it would be selling

Intel-based PCs preloaded with Lindows as an alternative to the now standard practice of

preloading PCs with Microsoft Windows. As discussed above, the youth of the computer

software market raised a variety of potential questions about whether the current state of

the market will be representative of the future of the market. Alternatives such as

Lindows, while likely insufficient for consumers today, present interesting alternatives

worth researching.


Works Cited and Consulted Allison, Juliann Emmons, ed. Technology, Development, and Democracy. Albany, NY: State

University of New York Press, 2002. Andrews, Paul. How The Web Was Won. New York: Broadway Books, 1999. “Apple – QuickTime Player.” http://www.apple.com/quicktime/products/qt/. Retrieved from the

World Wide Web 2003 April 21. Baker, C. Edwin. Advertising and a Democratic Press. Princeton, NJ: Princeton University

Press, 1994. Baker, C. Edwin. Human Liberty and Freedom of Speech. New York: Oxford University Press,

1989. Baker, C. Edwin. Media, Markets, and Democracy. New York: Cambridge University Press,

2002. Bank, David. Breaking Windows: How Bill Gates Fumbled the Future of Microsoft. New

York: The Free Press, 2001. Berg, Sanford V. and John Tschirhart. Natural Monopoly Regulation: Principles and Practice.

New York: Cambridge University Press, 1988. Dahl, Robert A. On Democracy. New Haven, CT: Yale University Press, 1998. Dahl, Robert A. A Preface to Economic Democracy. Berkeley, CA: The University of

California Press, 1985. “Defining the Basic Elements of .NET.” http://www.microsoft.net/net/basics/whatis.asp.

Retrieved from the World Wide Web 1 May 2003. Emord, Johnathan W. Freedom, Technology, and the First Amendment. San Francisco, CA:

Pacific Research Institute for Public Policy, 1991. Ferdinand, Peter, ed. The Internet, Democracy, and Democratization. Portland, OR: Frank Cass

Publishers, 2000. Frank, Mark W. The Impact of Rate-of-Return Regulation on Technological Innovation.

Burlington, VT: Ashgate Publishing Co., 2001. Heilemann, John. Pride Before the Fall: The Trials of Bill Gates and the End of the Microsoft

Era. New York: HarperCollins Publishers, Inc., 2001.

http://www.apple.com/quicktime/products/qt/

http://www.microsoft.net/net/basics/whatis.asp


Hoff, Jens, Ivan Horrocks and Pieter Tops, ed. Democratic Convergence and New Technology.

New York: Routledge, 2000. Inmon, William H. Technomics: The Economics of Technology and the Computer Industry.

Homewood, IL: Dow Jones-Irwin, 1986. John, Richard R. Spreading the News: The American Postal System from Franklin to Morse.

Cambridge, MA: Harvard University Press, 1995. Katsh, M. Ethan. The Electronic Media and the Transformation of Law. New York: Oxford

University Press, 1989. Kwoka, John E. Jr. and Lawrence J. White. The Antitrust Revolution: Economics, Competition,

and Policy. New York: Oxford University Press, 1999. Lin, Caroyln A. and David J. Atkin, ed. Communication Technology and Society: Audience

Adoption and Uses. Cresskill, NJ: Hampton Press, Inc., 2002. “Lindows OS.” http://www.lindows.com/lindows_sales_intro.php. Retrieved from the World

Wide Web 1 May 2003. Lowe, Janet. Bill Gates Speaks. New York: John Wiley & Sons, Inc., 1998. Mansell, Robin and W. Edward Steinmueller. Mobilizing the Information Society. Oxford:

Oxford University Press, 2000. Microsoft Corporation. Annual Report. Redmond, WA, Microsoft Corporation, 1986 – 2002.

(note: for brevity’s sake, this entry covers all the annual reports from 1986 – 2002 which I have acquired for my research)

Mill, John Stuart. On Liberty and Utilitarianism. New York, Random House, Inc., 1992. Morris, Irwin. Lecture and discussion. “GVPT170H: Introduction to American Government.”

University of Maryland: College Park, MD, Jan. 1999 – May 1999. “The New York Times on the Web.” http://www.nytimes.com/. Retrieved from the World Wide

Web 15 Mar 2003. Myers, David G. Social Psychology (6th ed.). New York: McGraw-Hill, 1999. Pennock, J. Roland and John W. Chapman, ed. Coercion. New York: Atherton, Inc., 1972. Reynolds, Alan. The Microsoft Antitrust Appeal. Westfield, IN: Hudson Institute, Inc., 2001.

http://www.lindows.com/lindows_sales_intro.php

http://www.nytimes.com/


Rohm, Wendy Goldman. The Microsoft File. New York: Times Books, 1998. “Seventh Annual BSA Global Software Piracy Study.” Washington, DC: Business Software

Alliance, June 2002. http://www.bsa.org/usa/policyres/admin/2002-06-10.130.pdf. (Electronic copy of the report) Retrieved from the World Wide Web 30 Apr 2003.

“Software Piracy Costs Businesses Billions.”

http://www.usatoday.com/tech/columnist/ccjoe026.htm. Retrieved from the World Wide Web 30 Apr 2003.

“Software Piracy Costs Billions.” http://www.wired.com/news/business/0,1367,13019,00.html.

Retrieved from the World Wide Web 30 Apr 2003. Shanor, Charles A. American Constitutional Law: Structure and Reconstruction. St. Paul, MN:

West Group, 2001. Silberschatz, Avi, Peter Galvin, and Greg Gagne. Applied Operating Systems Concepts (1st ed.).

New York: John Wiley & Sons, Inc., 2000. Smith, Robert Ellis. Ben Franklin’s Web Site: Privacy and Curiosity from Plymouth Rock to

the Internet. Providence, RI: Sheridan Books, 2000. Sunstein, Cass R. Democracy and the Problem of Free Speech. New York: The Free Press,

1993. Unger, Roberto Mangabeira. Knowledge and Politics. London: Collier Macmillian Publishers,

1975. Van Alstyne, William W. First Amendment: Cases and Materials (2nd ed.). Westbury, NY:

The Foundation Press, Inc., 1995. “The Wall Street Journal Online.” http://online.wsj.com/public/us. Retrieved from the World

Wide Web 15 Mar 2003. “WalMart.com to Sell Lindows OS.” http://www.idg.net/ic_1122987_9677_1-5042.html.

Retrieved from the World Wide Web 5 Feb 2003. “Windows XP: How To Buy.” http://www.microsoft.com/windowsxp/howtobuy/default.asp

Retrieved from the World Wide Web 2003 April 21. “Windows XP Professional Overview.”

http://www.microsoft.com/windowsxp/pro/evaluation/overviews/default.asp. Retrieved from the World Wide Web 2003 April 21.

http://www.bsa.org/usa/policyres/admin/2002-06-10.130.pdf

http://www.usatoday.com/tech/columnist/ccjoe026.htm

http://www.wired.com/news/business/0,1367,13019,00.html

http://online.wsj.com/public/us

http://www.idg.net/ic_1122987_9677_1-5042.html

http://www.microsoft.com/windowsxp/howtobuy/default.asp

http://www.microsoft.com/windowsxp/pro/evaluation/overviews/default.asp

Figure 1 – Microsoft Windows XP Network Setup Wizard

Figure 1.1 – First Wizard Screen

(This is what a user sees first after clicking on “Setup a New Connection” from the control panel or the automated setup that runs the first time Windows is started.)

Figure 1.2 – Second Wizard Screen

(This is what a user sees after clicking on the default option in 1.1 “Connect to the Internet”)

Figure 1.3 – Final Wizard Screen

(This is what the user sees after clicking on the default option “Choose from a list of Internet service providers (ISPs)” in figure 1.2 above. Selecting get online with MSN will automatically launch the MSN setup which comes

installed by default as part of Microsoft Windows XP.)

The Effects of Monopoly in Information Technology on...

Documents

Transcript of The Effects of Monopoly in Information Technology on...